To Err One’s Dirty Laundry* Gitte Larsen, MD, MPH Division of Pediatric Critical Care Primary Children’s Hospital University of Utah School of Medicine Salt Lake City, UT Jeff E. Schunk, MD Division of Pediatric Emergency Medicine University of Utah School of Medicine Salt Lake City, UT
I
n medicine, it is not infrequent for a colleague to start a conversation with a phrase akin to “do you remember that patient with [brief detail of patient history] that you saw last week?” The next part of the conversation may outline the superb care you delivered and your colleague’s undying respect for your clinical acumen, or, more likely, you will be given a brief synopsis of a diagnostic error (DE) that occurred. The error probably rose to this level of reporting if it was interesting and/or the outcome poor. However, this type of informal clinical feedback is dependent on complex interpersonal relationships within a medical community that are functional and respectful. More often colleagues do not provide feedback regarding errors. The obstacles are myriad and likely include an unequal power gradient, lack of familiarity with the individual, sympathy and understanding (e.g., anyone would miss this), fear of an uncomfortable conversation, assumption that the provider is already aware, and absence of an established system to identify DEs. For most physicians and nearly all diagnoses, there is no organized system that recreates this occasional chance collegial interaction. Medical providers receive little, if any, feedback regarding accuracy or timeliness of diagnoses or other medical errors. The concept of “misdiagnosis-related harm,” defined as preventable harm that results from the delay or failure to treat a condition actually present or from treatment provided for a condition not actually present (1), has been identified by autopsy, self-reported physician and patient surveys, malpractice claims, case reviews, and voluntary reports (2). Academic medicine has a long-standing tradition for providing feedback, often named morbidity and mortality conference (MMC). Case identification varies dramatically and could include feedback from other divisions or departments, patient complaints, poor outcomes (e.g., death and wound infection), *See also p. 468. Key Words: diagnostic error; mortality and morbidity conference; patient safety; pediatric intensive care unit; quality improvement The authors have disclosed that they do not have any potential conflicts of interest. Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000399
488
www.pccmjournal.org
system filters (e.g., cardiopulmonary resuscitation was performed), or other identifiers. In a previously published survey by Cifra et al (3) on U.S. PICU MMC, there was marked variability in adherence to a structured medical incident analysis, including a structured plan to investigate contributing factors to adverse events. The authors point out that one source of variability is likely the evolution of MMC from an educational endeavor to a more active role in improving patient safety (3), although other authors have suggested the goals since conception have been both education and system improvement (4). Effectiveness and utility of MMC have not been well studied, although preliminary studies suggest that MMCs are better at identifying near misses, communication/teamwork errors, workflow issues, and other broader issues not easily identified by chart review (3). In this issue of Pediatric Critical Care Medicine, Cifra et al (5) describe DE identified from a 96-PICU patient cohort over a 14-month period. Although this is a single institution investigation, it provides valuable insight into DE. The authors should be congratulated for sharing their errors and using a rigorous analysis. Key points from this article include DE identification and categorization. The majority were self-reported in the MMC process (55%); however, 35% were discovered at autopsy. The authors felt that 95% of the DE could have had an impact on safety or survival. DE cannot be reliably detected by chart review, as the record rarely reflects the nuances of provider thinking and medical decision making (5). The record more commonly outlines management already performed and results of investigations; there are only cursory discussions of diagnostic possibilities or considerations. Chart review has been found useful when identifying delays in diagnosis for pediatric trauma patients, partly due to a more straightforward labeling system for injuries (6). For the PICU setting, DE most commonly occurred within vascular and neurologic disease categories (5). The authors modified a classification system originally designed for autopsy findings, to rate the severity of DE in their patient cohort, many of whom did not die (5). This is an important step toward error classification, and future studies should determine generalizability and utility. The authors appropriately present limitations to this study, including the high likelihood of underreporting PICU DE; the sample included only those cases discussed at MMC (5). In general, DE reporting likely needs to meet two thresholds. 1) Recognition: as rudimentary as this sounds, someone needs to realize that we are now managing a patient with condition X, previously managed as condition Y. 2) Severity or interest: DE needs to be significant, off the mark (it had not been previously considered), or interesting. Established criteria for DE and diagnostic delay do not exist; the authors have employed MMC with structured, independent, and tiered reviews to enhance identification of DE (5). The findings suggest that DE occurs in the PICU more frequently than in the care delivered prior to the PICU; however, the methodology did not lend itself to June 2015 • Volume 16 • Number 5
Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies. Unauthorized reproduction of this article is prohibited
Editorials
generating this conclusion (5). Furthermore, advanced testing or timing of test results may be delayed until the patient is in the PICU, so it is difficult to establish the accuracy of an earlier diagnosis. The importance of chronology or disease natural history in the discovery of DE requires emphasis. Clinicians work within a diagnostic paradigm, and there is an inherent expectation of treatment effect with time (e.g., the cellulitis will improve with 24 hr of IV antibiotics) or typical time course for improvement or resolution (e.g., the weakness in the left arm should resolve within a few hours following the focal seizure). When these expectations are not met, then we search for an alternative explanation (e.g., our organism assumption was wrong, this is the wrong antibiotic) or diagnosis (e.g., this could be a stroke). In this respect, DE discovery is tied to timing and chronology of disease characteristics. Future investigations will need to address what could be acceptable standards for delaying a diagnosis within this framework. Although some errors can be identified at a system level (e.g., medication dosing in computerized electronic order entry) or via the use of a “global trigger tool” (7), detecting errors in diagnosis are particularly challenging. Potential approaches include using case reviews with a “trigger tool” to
identify high-risk cases for DE, analyzing autopsy results, and encouraging voluntary reporting on the part of patients and care team members (2). DEs are likely to remain unreported or difficult to recognize until the science of measurement, including classification, improves (1).
REFERENCES
1. Newman-Toker DE, Pronovost PJ: Diagnostic errors—The next frontier for patient safety. JAMA 2009; 301:1060–1062 2. Graber ML: The incidence of diagnostic error in medicine. BMJ Qual Saf 2013; 22(Suppl 2):ii21–ii27 3. Cifra CL, Bembea MM, Fackler JC, et al: The morbidity and mortality conference in PICUs in the United States: A national survey. Crit Care Med 2014; 42:2252–2257 4. Deshpande JK, Throop PG, Slayton JM: Standardization of case reviews (morbidity and mortality rounds) promotes patient safety. Pediatr Clin North Am 2012; 59:1307–1315 5. Cifra CL, Jones KL, Ascenzi JA, et al: Diagnostic Errors in a PICU: Insights From the Morbidity and Mortality Conference. Pediatr Crit Care Med 2015; 16:468–476 6. Furnival RA, Woodward GA, Schunk JE: Delayed diagnosis of injury in pediatric trauma. Pediatrics 1996; 98:56–63 7. Agarwal S, Classen D, Larsen G, et al: Prevalence of adverse events in pediatric intensive care units in the United States. Pediatr Crit Care Med 2010; 11:568–578
Another Step in Understanding Glucocorticoid Resistance in Critical Illness* Ricardo Garcia Branco, MD, PhD Paediatric Intensive Care Unit Cambridge University Hospitals NHS Trust Cambridge, United Kingdom Pedro Celiny Ramos Garcia, MD, PhD Department of Paediatrics Hospital Sao Lucas da PUCRS Porto Alegre, Brazil Robert Charles Tasker, MBBS, MD Departments of Neurology and Anesthesia Boston Children’s Hospital and Harvard Medical School Boston, MA
*See also p. e132. Key Words: corticosteroids; glucocorticoid receptor; glucocorticoid resistance; pediatric critical care; sepsis Dr. Garcia received grant support from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior/Conselho Nacional de Desenvolvimento Científico e pesquisa (Brazilian National Research Grants). Dr. Tasker received royalties from the Oxford Handbook of Paediatrics. Dr. Branco has disclosed that he does not have any potential conflicts of interest. Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000421
Pediatric Critical Care Medicine
O
ver 60 years ago physicians tried to modulate the body’s response to severe infection using corticosteroids (1). This intervention was the beginning of a now long-running debate—What is the role of corticosteroid supplementation in severe illness? Over recent decades, thousands of studies have significantly improved our understanding of hypothalamic-pituitary-adrenal (HPA) axis (dys) function during critical illness, but a definitive role for corticosteroid therapy has not been demonstrated. The contemporary description of the imbalance between a patient severity of illness and their corticosteroid activity is defined as critical illness–related corticosteroid insufficiency (CIRCI) (2). This term encompasses a wide range of physiological changes that may or may not be adaptive but results in reduced corticosteroid activity that is considered inappropriate during critically illness. For example, a patient may have an inappropriate cortisol level in relation to their illness or may have adequate cortisol levels but an inadequate reserve for mounting any increase in cortisol levels if required. These two situations have been adequately studied (3, 4), and so far, investigators have failed to identify children who benefit from corticosteroid therapy. In contrast, CIRCI is a far more complex syndrome, where corticosteroid activity may be impaired by a number of mechanisms, including central dysregulation of the HPA axis, inadequate cortisol production or metabolism, www.pccmjournal.org
489
Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies. Unauthorized reproduction of this article is prohibited
I
n medicine, it is not infrequent for a colleague to start a conversation with a phrase akin to “do you remember that patient with [brief detail of patient history] that you saw last week?” The next part of the conversation may outline the superb care you delivered and your colleague’s undying respect for your clinical acumen, or, more likely, you will be given a brief synopsis of a diagnostic error (DE) that occurred. The error probably rose to this level of reporting if it was interesting and/or the outcome poor. However, this type of informal clinical feedback is dependent on complex interpersonal relationships within a medical community that are functional and respectful. More often colleagues do not provide feedback regarding errors. The obstacles are myriad and likely include an unequal power gradient, lack of familiarity with the individual, sympathy and understanding (e.g., anyone would miss this), fear of an uncomfortable conversation, assumption that the provider is already aware, and absence of an established system to identify DEs. For most physicians and nearly all diagnoses, there is no organized system that recreates this occasional chance collegial interaction. Medical providers receive little, if any, feedback regarding accuracy or timeliness of diagnoses or other medical errors. The concept of “misdiagnosis-related harm,” defined as preventable harm that results from the delay or failure to treat a condition actually present or from treatment provided for a condition not actually present (1), has been identified by autopsy, self-reported physician and patient surveys, malpractice claims, case reviews, and voluntary reports (2). Academic medicine has a long-standing tradition for providing feedback, often named morbidity and mortality conference (MMC). Case identification varies dramatically and could include feedback from other divisions or departments, patient complaints, poor outcomes (e.g., death and wound infection), *See also p. 468. Key Words: diagnostic error; mortality and morbidity conference; patient safety; pediatric intensive care unit; quality improvement The authors have disclosed that they do not have any potential conflicts of interest. Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000399
488
www.pccmjournal.org
system filters (e.g., cardiopulmonary resuscitation was performed), or other identifiers. In a previously published survey by Cifra et al (3) on U.S. PICU MMC, there was marked variability in adherence to a structured medical incident analysis, including a structured plan to investigate contributing factors to adverse events. The authors point out that one source of variability is likely the evolution of MMC from an educational endeavor to a more active role in improving patient safety (3), although other authors have suggested the goals since conception have been both education and system improvement (4). Effectiveness and utility of MMC have not been well studied, although preliminary studies suggest that MMCs are better at identifying near misses, communication/teamwork errors, workflow issues, and other broader issues not easily identified by chart review (3). In this issue of Pediatric Critical Care Medicine, Cifra et al (5) describe DE identified from a 96-PICU patient cohort over a 14-month period. Although this is a single institution investigation, it provides valuable insight into DE. The authors should be congratulated for sharing their errors and using a rigorous analysis. Key points from this article include DE identification and categorization. The majority were self-reported in the MMC process (55%); however, 35% were discovered at autopsy. The authors felt that 95% of the DE could have had an impact on safety or survival. DE cannot be reliably detected by chart review, as the record rarely reflects the nuances of provider thinking and medical decision making (5). The record more commonly outlines management already performed and results of investigations; there are only cursory discussions of diagnostic possibilities or considerations. Chart review has been found useful when identifying delays in diagnosis for pediatric trauma patients, partly due to a more straightforward labeling system for injuries (6). For the PICU setting, DE most commonly occurred within vascular and neurologic disease categories (5). The authors modified a classification system originally designed for autopsy findings, to rate the severity of DE in their patient cohort, many of whom did not die (5). This is an important step toward error classification, and future studies should determine generalizability and utility. The authors appropriately present limitations to this study, including the high likelihood of underreporting PICU DE; the sample included only those cases discussed at MMC (5). In general, DE reporting likely needs to meet two thresholds. 1) Recognition: as rudimentary as this sounds, someone needs to realize that we are now managing a patient with condition X, previously managed as condition Y. 2) Severity or interest: DE needs to be significant, off the mark (it had not been previously considered), or interesting. Established criteria for DE and diagnostic delay do not exist; the authors have employed MMC with structured, independent, and tiered reviews to enhance identification of DE (5). The findings suggest that DE occurs in the PICU more frequently than in the care delivered prior to the PICU; however, the methodology did not lend itself to June 2015 • Volume 16 • Number 5
Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies. Unauthorized reproduction of this article is prohibited
Editorials
generating this conclusion (5). Furthermore, advanced testing or timing of test results may be delayed until the patient is in the PICU, so it is difficult to establish the accuracy of an earlier diagnosis. The importance of chronology or disease natural history in the discovery of DE requires emphasis. Clinicians work within a diagnostic paradigm, and there is an inherent expectation of treatment effect with time (e.g., the cellulitis will improve with 24 hr of IV antibiotics) or typical time course for improvement or resolution (e.g., the weakness in the left arm should resolve within a few hours following the focal seizure). When these expectations are not met, then we search for an alternative explanation (e.g., our organism assumption was wrong, this is the wrong antibiotic) or diagnosis (e.g., this could be a stroke). In this respect, DE discovery is tied to timing and chronology of disease characteristics. Future investigations will need to address what could be acceptable standards for delaying a diagnosis within this framework. Although some errors can be identified at a system level (e.g., medication dosing in computerized electronic order entry) or via the use of a “global trigger tool” (7), detecting errors in diagnosis are particularly challenging. Potential approaches include using case reviews with a “trigger tool” to
identify high-risk cases for DE, analyzing autopsy results, and encouraging voluntary reporting on the part of patients and care team members (2). DEs are likely to remain unreported or difficult to recognize until the science of measurement, including classification, improves (1).
REFERENCES
1. Newman-Toker DE, Pronovost PJ: Diagnostic errors—The next frontier for patient safety. JAMA 2009; 301:1060–1062 2. Graber ML: The incidence of diagnostic error in medicine. BMJ Qual Saf 2013; 22(Suppl 2):ii21–ii27 3. Cifra CL, Bembea MM, Fackler JC, et al: The morbidity and mortality conference in PICUs in the United States: A national survey. Crit Care Med 2014; 42:2252–2257 4. Deshpande JK, Throop PG, Slayton JM: Standardization of case reviews (morbidity and mortality rounds) promotes patient safety. Pediatr Clin North Am 2012; 59:1307–1315 5. Cifra CL, Jones KL, Ascenzi JA, et al: Diagnostic Errors in a PICU: Insights From the Morbidity and Mortality Conference. Pediatr Crit Care Med 2015; 16:468–476 6. Furnival RA, Woodward GA, Schunk JE: Delayed diagnosis of injury in pediatric trauma. Pediatrics 1996; 98:56–63 7. Agarwal S, Classen D, Larsen G, et al: Prevalence of adverse events in pediatric intensive care units in the United States. Pediatr Crit Care Med 2010; 11:568–578
Another Step in Understanding Glucocorticoid Resistance in Critical Illness* Ricardo Garcia Branco, MD, PhD Paediatric Intensive Care Unit Cambridge University Hospitals NHS Trust Cambridge, United Kingdom Pedro Celiny Ramos Garcia, MD, PhD Department of Paediatrics Hospital Sao Lucas da PUCRS Porto Alegre, Brazil Robert Charles Tasker, MBBS, MD Departments of Neurology and Anesthesia Boston Children’s Hospital and Harvard Medical School Boston, MA
*See also p. e132. Key Words: corticosteroids; glucocorticoid receptor; glucocorticoid resistance; pediatric critical care; sepsis Dr. Garcia received grant support from Coordenação de Aperfeiçoamento de Pessoal de Nível Superior/Conselho Nacional de Desenvolvimento Científico e pesquisa (Brazilian National Research Grants). Dr. Tasker received royalties from the Oxford Handbook of Paediatrics. Dr. Branco has disclosed that he does not have any potential conflicts of interest. Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies DOI: 10.1097/PCC.0000000000000421
Pediatric Critical Care Medicine
O
ver 60 years ago physicians tried to modulate the body’s response to severe infection using corticosteroids (1). This intervention was the beginning of a now long-running debate—What is the role of corticosteroid supplementation in severe illness? Over recent decades, thousands of studies have significantly improved our understanding of hypothalamic-pituitary-adrenal (HPA) axis (dys) function during critical illness, but a definitive role for corticosteroid therapy has not been demonstrated. The contemporary description of the imbalance between a patient severity of illness and their corticosteroid activity is defined as critical illness–related corticosteroid insufficiency (CIRCI) (2). This term encompasses a wide range of physiological changes that may or may not be adaptive but results in reduced corticosteroid activity that is considered inappropriate during critically illness. For example, a patient may have an inappropriate cortisol level in relation to their illness or may have adequate cortisol levels but an inadequate reserve for mounting any increase in cortisol levels if required. These two situations have been adequately studied (3, 4), and so far, investigators have failed to identify children who benefit from corticosteroid therapy. In contrast, CIRCI is a far more complex syndrome, where corticosteroid activity may be impaired by a number of mechanisms, including central dysregulation of the HPA axis, inadequate cortisol production or metabolism, www.pccmjournal.org
489
Copyright © 2015 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies. Unauthorized reproduction of this article is prohibited
